1. Field of the Invention
This invention relates to spectrochemical determination of analyte content of a sample, and more particularly to a spectroscopy sample cell for analyte analysis of turbid media such as blood.
2. Background Information
Spectrochemical analysis is a broad field in which the composition and properties of materials in any phase, (liquids, solids, gases or plasmas) are determined from the electromagnetic spectra arising from interaction (for example, absorption, luminescence or emission) with energy. One aspect of spectrochemical analysis, known as spectroscopy, involves interaction of radiant energy with material of interest. The particular methods used to study such matter-radiation interactions define many sub-fields of spectroscopy. One field in particular is known as absorption spectroscopy, in which the optical absorption spectra of liquid substances is measured. The absorption spectra is the distribution of light attenuation (due to absorbance) as a function of light wavelength. In a simple spectrophotometer, the sample substance which is to be studied, usually a gas or liquid, is placed into a transparent container, also known as a cuvette or sample cell. Collimated light of a known wavelength, .lambda., i.e. ultraviolet, infrared, visible, etc., and intensity I.sub.o, is incident on one side of the cuvette. A detector, which measures the intensity of the exiting light, I, is placed on the opposite side of the cuvette. The thickness of the sample is the distance d, that the light propagates through the sample. For a sample consisting of a single, homogenous substance with a concentration, c, the light transmitted through the sample will follow a well known relationship known as Beer's law which in part, states that the shorter the optical pathlength d, the less the absorption of light. If .lambda., I.sub.o, I, d, c, are known, a quantity known as the extinction coefficient can be determined. EQU OD(.lambda.)=c*.epsilon.(.lambda.)*d
Where:
.epsilon.(.lambda.) is the extinction coefficient of the substance at wavelength .lambda.; PA0 OD(.lambda.) is the optical density of the sample at wavelength .lambda. (OD=-Log of the ratio: light transmitted (exiting)/light incident. (-Log.sub.10 I/I.sub.0) PA0 c is the concentration of the substance PA0 d is the pathlength or thickness of the sample through which the light propagates.
Beer's law is useful when one considers samples which are mixtures of several different substances, j, each with known extinction coefficients, .epsilon..sub.j, and relative concentrations c.sub.j. In such a case, the optical density of the sample is given by: EQU OD(.lambda.)=.SIGMA.c.sub.j *.epsilon..sub.j *(.lambda.)*d
Inversely, if given a sample mixture's optical density spectra OD(.lambda.), and the extinction coefficients for each of the component substances, the unknown relative concentration of the component substances can be determined. Note, also, it is assumed that each of the component substances maintain the same extinction coefficients as when in a pure form, i.e. no chemical reactions occur that alter the extinction coefficients.
Thus, if the absorption spectra for a given substance is known, its presence and concentration in a sample may be determined.
Optimum optical pathlength depends on a variety of factors. In general, relatively small pathlengths are preferred as they enable use of smaller sample volumes and facilitate optical measurement of samples, such as blood, which have relatively high optical density, absorbance or light scattering properties. However, problems tend to occur when pathlengths as small as a few thousandths of an inch are used with whole blood samples. Unfiltered human blood may contain various formed elements (cell aggregates, protein/fibrin chains, synthetic fibers, clots, etc.), which when introduced into a narrow sample cell, can become lodged in the sample cell. These "clots" can obstruct the flow of wash fluids or subsequent samples, which can adversely affect the optical measurements. When high velocity air or liquid is forced through the sample cell, these clots can sometimes be dislodged. The introduction of a cleaning solution containing bleach or pepsin may also be effective in removing protein and fibrin from the sample cell. These cleaning solutions however, can only be used in systems which can tolerate exposure to such agents. In some cases, the sample cell may have to be replaced or disassembled and manually cleaned to remove such clots.
Optical measurements in whole blood can be further complicated by a number of issues. When red blood cells, the small "carriers" of hemoglobin (the light absorbing constituent to be measured), are suspending in serum, a light scattering (non isotropic) medium is produced. This is due to the differing indices of refraction of the red blood cell's intracellular fluid and the serum in which the cells are suspended. In such a case, Beer's law may not strictly apply and light scattering in the sample may result in significant errors in spectrophotometric measurements.
The degree of light scattering in whole blood is influenced by a number of normally occurring conditions such as variations in serum protein and lipid content, red cell morphology (size and shape), red cell concentration (number of red cells per unit volume), cell rouleaux formation, and red cell orientation.
Rouleaux formation occurs when a sample of whole blood is allowed to remain stationary, whereupon the red blood cells coalesce in an ordered fashion, forming "stacks" or "chains." This phenomenon, which is due, in part, to the bi-concave shape of the red cells and the colloids present in blood serum, significantly affects the optical properties of whole blood. Although some exceptions exist (e.g., certain animal bloods and rare variations in cell morphology/serum chemistry), this phenomenon occurs normally in most whole blood samples. The rate at which these rouleaux "chains" form is strongly dependent on red cell concentration.
Rouleaux formation is a reversible phenomenon. If the blood sample is mixed, stirred or made to flow through a channel of some type, resulting shear forces within the fluid will cause the rouleaux chains to disassemble and break up. Stopping the blood flow will allow, once again, the chains to form. If allowed to sit for extended periods of time, these rouleaux chains will settle and eventually result in the separation (stratification) of the blood cells and the blood serum.
Rouleaux formation alters the amount of light scatter observed in a given blood sample. It is important then, that a means of either accounting for or preventing the effect must be implemented. One way of preventing rouleaux chains from forming, is to force the blood to flow through a small channel. Shear forces, generated by differential velocities (velocity gradients) throughout the flow channel, if high enough, tend to prevent rouleaux chains from forming. A drawback of this approach however, is that use of a relatively small channel disadvantageously exacerbates the tendency for clogging or clotting, etc.
Another factor to be considered is whether or not the red cells are "oriented" during analyte measurement. Red cells tend to become oriented in a particular manner while flowing through narrow channels at relatively high velocity. A "normal" red blood cell is bi-concave (i.e. shaped much like a donut; thinner in the middle than at the edges) and measures approximately 8 microns (8.times.10.sup.-6 meters) across by 2 microns thick. A single normal red blood cell, if suspended in a carrier fluid experiencing relatively high velocity laminar flow, will be acted upon by the shear forces generated by the differential velocities in the fluid. Because of the cell's asymmetric shape, these forces will tend to orient the cells in some non-random fashion. This orientation of red blood cells occurs in mass (whole blood) flow, and is affected by the size and shape of the flow cell, particularly by use of flow cells having relatively small or restricted openings.
The amount of light transmission through a sample of whole blood where the red cells are randomly organized differs significantly from one in which the cells are oriented (i.e. non-randomly). This aspect of the red cells thus tends to further complicate spectroscopic analyte measurement in whole blood samples.
One attempt to overcome the aforementioned difficulties and provide accurate optical determination of analytes in blood samples has been embodied in instrumentation manufactured by Bayer Corporation of Medfield, Mass., known as the 800 Series analyzer. This instrumentation performs spectrophotometric measurements using lysed blood samples. The lysing process ruptures the red blood cell's membrane, releasing the intracellular fluid (primarily hemoglobin) into the surrounding serum. Lysing the red cells makes the sample relatively homogeneous and isotropic (i.e. essentially non-light scattering). Beer's law, as discussed above, then may be effectively applied to determine analyte concentration.
Ultrasonic energy provided by ultrasonic cell disrupters is typically utilized to lyse the red cells. A drawback of this approach is that these cell disrupters add cost, take up space, require maintenance and may generate bubbles in the sample which generate unpredictable light scatter induced errors. Also, the lysing process alters the sample's serum chemistry. Because of this, integrated instruments which combine lysed blood spectrophotometry with other sensor technologies that require whole blood for their measurements require additional sample volume for analysis. In these systems, a sample is aspirated into the instrument, then "split" into two segments to effectively isolate a portion of the sample for lysing and an other portion for whole blood analysis.
Thus, a need exists for an improved spectroscopy sample cell that avoids clogging, may be easily cleaned without special cleaning agents, is optimized for various operating modes including measurement, wash, fill, etc., is inexpensive and provides a stable mechanical environment for optical measurement repeatability.